Bearings

dave williams
rec.autos.tech  08-27-93
- Interestingly, back in the 1920s and 1930s, roller lifters and
  rockers were quite common, particularly on better-brand cars.  Tubular
  manifolds weren't unknown either.
- The rollers were used because the lubricants of the day weren't good
  enough to provide the longevity the designers wanted.  Once high
  pressure additives became available, designers specified better oil and
  went to the cheaper flat tappets.

[email protected] (Dave Williams)
gnttype  28 Jan 1994
- -> There was an article in the GSXtra a couple of issues ago discussing
  -> this for the 455 motors. The basic conclusion was that it was better
  -> to cross drill than to loose the bearing surface.
- Correct.  If one extra hole is going to make a difference in crank
  strength, the crank was too weak to begin with.
- Grooved bearings are a problem in that a sleeve bearing is more
  dependent on width than diameter.  When you groove the bearing, you have
  two very narrow bearings instead of one wider one, and the load capacity
  goes down drastically.

[email protected] (Dave Williams)
gnttype  28 Jan 1994
- -> Do you really lose load capacity with grooved bearings?
- Yes, and the amount is substantial.  Do you have a Machinery's
  Handbook?  If not, the library might.  The section on plain bearings is
  quite interesting.
- Basically, the load isn't distributed evenly across the bearing.  The
  center carries most of the load, with the load dropping off at the
  edges, where oil is bled away.
- -> The old GS455s apparently used fully grooved bearings.
- Lots of engines did and do.  That's because the connecting rods are
  loaded more than the crank, and they're running on narrower, smaller
  diameter bearings too!  The designers were trading some main bearing
  load for better lubrication for the rods.
- -> Would it improve crank strength to drill it at 120 degree spacings,
  -> i.e. three holes, rather than 180? That way you wouldn't have a
  -> straight "parting line" across the crank. I wonder if anybody has
  -> heard/tried this out.
- I've never come across the idea, but it's interesting.  With a grooved
  upper shell, you have 180 degrees of oil feed.  Technically a straight
  drilled crank would feed oil to the rods at all times.  In actuality,
  the oil has to start and stop, and it has a fairly long travel path from
  the oil feed point at the shell, through the groove, and around to the
  side of the crank where the oil hole is.  The hole is moving, which
  ain't the best way to ensure steady flow, either.
- With the crossdrilled crank, you never have a hole more than 45 degrees
  from the feed, as opposed to 180 max for the single drilled crank.  I'd
  tend to guess that, if you're having oiling related problems with a
  cross drilled crank, a three-drilled crank wouldn't help you any.

[email protected] (Dave Williams)
gnttype  03 Mar 1994
- -> surface and corrosion characteristics.  I think the standard for
  -> years were the all aluminum alloy type or the babbit coated ones.  Go
  -> the few extra bucks and get the most up to date in journal bearing
  -> technology.
- The aluminum bearings aren't seen much.  Some heavy duty truck motors
  use them, and Diesels, but the only aluminum you see in a normal car
  bearing is aluminum-bronze.
- Chrysler kept with Babbitt bearings in some locations all the way up to
  the '60s, but the rest of the industry had changed over to trimetal in
  the '50s.  Plain bearing design hasn't changed significantly since.  You
  can futz around with the hardnesses of the layers to trade embeddability
  for strength, but that's about it.
- -> Incidentally, for journal bearing prep, I think the moly EP grease is
  -> best
- I used to do that, but Chevrolet's Chevy Power manual says not to, as
  it will stop up the oil filter.  I switched to STP for the bearings
  after that, and the bottoms of the lifters, the rocker wear areas, etc.
  Everything else gets a squirt from a pump oiler filled with Mobil 1 synthetic.
- -> Most of it dripped off the bearings.  The moly is what is in GM's
  -> E.O.S. and this is supposedly good prep.
- GM EOS (and Ford's equivalent) have lots of zinc in them as a high
  pressure additive.  If you're reringing the motor GM says not to use it
  on cars equipped with oxygen sensors.  I suspect some cam lubes are the
  same basic stuff.  A few hundred miles probably won't kill the sensor,
  though.
- -> this is supposedly good prep.  Heard you should change the oil/filter
  -> after lik 1 hour though because the moly clogs filters fast.
- That's about how I do mine.  The first hour of operation is the worst
  the engine is ever going to see.  That's when any debris is going to get
  circulated around, the cylinder walls are being lapped by the rings, and
  all sorts of monkey-motion is going on.  $10 for five quarts of oil and
  a filter is pretty damned cheap insurance if you ask me.
- -> Is there a way to prime the oiling system in the 3.8 before you crank
  -> it over the first time?
- Someone probably sells a pre-oiler, but cranking it over with the plugs
  out until oil pressure comes up works okay.  Not that I'd turn up my
  nose at a pre-oiler if someone offered me one, though.

chaos.lrk.ar.us!dave.williams (Dave Williams)
hotrod  10 Jul 1994
- A couple of years ago discussion touched on 360 degree oiling - instead
  of running grooved mains on both top and bottom, use a plain bottom
  bearing, cut an oiling groove in the block and cap, and drill holes as
  desired in the bottom bearing.  John put his hairy moderator Foot on
  the thread, something about "don't bring that one up again."  C'mon, John.
- [Moi?  I plead innocent due to loss of memory....  JGD]
- I built one engine this way, no trouble, just experimenting.
  Chevrolet's race book talks about it.  That's about all I've seen, so I
  missed out on whatever flamewars took place.  
- Fully-grooved mains (or grooved cranks, now considered a Very Bad Idea)
  were designed to get more oil up to the rod bearings.  Most cranks are
  drilled diagonally from the main journal up to the rod journal.  When
  the main journal hole is in the grooved area, oil (hopefully) flows
  through the drilling to the rod bearing.  Using a fully grooved bearing
  lets oil flow all the time, at the expense of losing well over half of
  the mains' load carrying capability.  Most engines use plain bearings on
  the bottom, grooved on the top, and the rods appear to live OK.  When
  there's question, the crank is usually cross-drilled.  A second hole is
  drilled, from the original rod oiling hole in the main journal, straight
  across to the other side of the crank.  When the crank turns so as to
  block oil flow to one side of the drilling, the other side comes up and
  starts taking oil.
- Now, that plain bottom bearing is only lubricated by what oil has
  managed to stay attached to the crank.  A few small drilled oil holes
  can feed more oil from the bottom, assuming you drill them on the low
  pressure side.  As the crank rotates it builds up a "wedge" of high
  pressure oil adjacent to the point of maximum load.  You don't want to
  put any oil holes there, or you might wind up with the oil wedge
  actually pumping oil right back into the oil galleries.
- Now, if the crank were just a smoothly spinning shaft, oiling from the
  bottom wouldn't be much help.  Modern theory is that, under load, the
  crank is writhing about like a snake, or as much as it can subject to
  the constraints of the bearing clearances.  So the nice oil flow and
  wedge model breaks down when the crank starts to whip and the wedge
  moves to the top, or the side, or wherever it wants.  So 360 degree
  oiling would (theoretically) ensure the crank was always getting some
  oil, no matter what was going on.
- Hank the Crank (are they still around?) and Bruce Crower favored
  putting *all* the oil in through the main cap, though I've yet to figure
  a real good reason for it.

[email protected] (Dave Williams)
gnttype  09 Dec 1995
- -> He showed me the bearings he had out and there was black "SCARRING"
  -> towards the center of all of them,
- Hmm.  That's not anything I've seen before.  Most modern bearings aren't 
  completely circular for various reasons, so it's common to see 1/4 to 3/8 
  inch of dark band at the edges where the top and bottom half touch each 
  other.  Worn areas at the center of the bearing would be coppery or silvery, 
  depending on how many layers of the bearing had been worn. 
- If the engine had been run low on oil or detonated a lot, you might see a 
  darkish area at around 10 or 2 o'clock, on the top shell only.  This is 
  burned oil, and though you'd ordinarily replace the bearing since you're 
  that far into the motor, it hasn't failed yet. 

[email protected] (Dave Williams)
fordnatics  09 Aug 1996
- -> can you explain that a bit?  does the heat buildup cause the bearings
  -> to expand and press too hard?
- Uh-oh, that's starting to get complex.  Stripped to the bare essentials, the 
  a thin layer of oil "sticks" to the crank and the bearing shell.  In between 
  is more oil which is sliding on the oil stuck to the first two layers. 
- The problem is, oil has viscosity, otherwise defined as "internal friction".  
  The molecules in motor oil are long and stringy, which is good for 
  lubrication, but the long molecules make a lot of heat sliding around each 
  other.  They also break through plain old mechanical shear -oil *does* "wear 
  out" despite the claims of miracle filter vendors. 
- The higher the loading on the bearing, the more heat the oil makes. 
  Eventually the bulk oil in the pan gets hot too.  Finally it will get hot 
  enough so the temperature rise in the bearing is enough to cause the oil to 
  break down very rapidly, char, or even reach flash point. However, before 
  this happens the oil is actually hot enough to start melting the soft 
  embeddable surface of the bearing, causing smearing, pits, or a spalling. 
- The characteristic of the 302 seems to be, the Ford bearings have lowish 
  melting point and get wiped out fairly easily.  Clearances open up, which 
  causes more oil to be pumped through, which reduces the temperatures, so 
  things appear to reach equilibrium somewhere about the time you get to the 
  copper plating in the bearing shell. 
- Practically any amount of oil will do for lubrication.  The reason you pump 
  mass quantities through at 40 to 100psi is for cooling the bearings.  And 
  since the oil runs pretty hot already, you have to pump even more of it to 
  get the same cooling effect.  You could probably chop pump output 'way back 
  if you could ensure a supply of room-temperature oil; in practice you're 
  hard pressed to keep it down below 220F in a competition engine. 
- If you have big, wide bearings, low RPM, and a low power output, you can get 
  away with splash oiling.  Pre'52 Chevy 216 six cylinders and Briggs&Stratton 
  lawnmower engines are examples.  They have a hole in the bottom of the rod 
  and a scoop to direct oil to the rod journal.  For the speeds and loads they 
  operate at, it works.  Many cars made before WWII had pressure lubrication 
  to the mains and splash to the rods. 
- Bear in mind this is the Readers' Digest Condensed Version. Lubricant design 
  used to be a branch of chemistry; now it's a subset of a distinct 
  engineering specialty known as tribology. 

[email protected] (Dave Williams)
fordnatics  27 Jan 1998
- -> I thought the reason for feeding the bearings from the bottom is
  -> because that i the load rests.  I know they are different but all our
  -> turbine and generator be fed from the bottom.
- When a shaft turns inside a plain bearing you get a funky seashell-shaped 
  pressure curve if you look at a polar plot.  That is, the pressure across 
  the bearing is uneven, with the highest point of pressure being about thirty 
  degrees from the load vector.  Leading that high point is the low pressure 
  point, where it is most advantageous to feed the bearing. 
- In an ideal situation you would indeed feed the bearing from the bottom, 
  just as it is in the Ford camshaft journals.  The sum of the load vectors 
  from the lifters is straight down.  Unfortunately most automobile bearings 
  are out there in the "special case" category.  They are *very* narrow for 
  plain bearings.  Plain bearings are sensitive to edge effects - pressure is 
  lost to the sides.  The aspect ratio of all the bearings in a car engine is 
  so low you are dealing *primarily* with edge effects.  So what happens is, 
  when they inject the oil to the bottom of the bearing, the bearing carries 
  it along fifteen degrees or so of rotation and then promptly spits it out 
  the sides rather than building the "hydrostatic wedge" which is what makes a 
  plain bearing work. 
- Thus, it works fine for your turbines, which presumably have properly 
  designed bearings.  But the narrow bearings in a car present a special case. 

[email protected] (Dave Williams)
diy_efi  06 May 1998
- -> regenerative braking and massive power switching module.  On top of
  -> all that, pure water was used for cooling of motor and get this: was
  -> fluid used in turbine's journal bearings instead of oil.
- Sure, water will work just fine.  Take a look at a big hydroelectric turbine 
  sometime; chances are it'll have water plumbed to sleeve bearings.  Once the 
  shaft is spinning any fluid will do - even air, which is often used in very 
  high speed applications. 
- Automobiles use oil-fed plain bearings primarily because they already *have* 
  oil in the crankcase to lube the myriad things which require lubrication, so 
  it makes sense to use oil. 

[email protected] (Dave Williams)
fordnatics  27 Dec 1998
- -> Anyone know why no one uses a desmodromic arraingement like Ducati?
  -> No springs at all. Cam opens and closes the valve.
  -> No way the valve can float then.
- I've considered this one myself, with no clear answer.  Even Ducati was 
  turning away from desmo last I heard. 
- Production costs would probably come out close to the same in OEM-volume 
  quantities.  I suspect it's mostly inertia - coil sprung valves are simple, 
  well understood, and do an adequate job with few failures. 
- I have been predicting a resurgence of desmodromic valvetrains for the last 
  few years, though.  Springs suck some power and cause friction; as 
  automakers writhe under the screws of emissions and fuel mileage 
  legislation, desmo valves will surely return, like other "obsolete" 
  technologies - four valve heads (over a century old now), coil-on-plug 
  (Fordson tractors used them before WWII), multiple spark ignition (Model T 
  Ford) etc. 
- We'll probably also see more use of ball and roller bearings.  Plain 
  bearings are more efficient than roller bearings, *but* they require 
  pressure fed oil.  Unfortunately a plain bearing's lubricant flow 
  requirements are opposite the way oil pumps work - they need more flow at 
  lower RPM, less at high.  Oil pumps are sized to give full flow at idle; 
  past that they dump excess oil overboard, heating and aerating it. We're 
  talking 5-15hp here, plus the benefits for engines like the 4.6 Ford, which 
  have a horrible oiling system to start with.  Cam journal wear is the #1 
  warranty problem on the 4.6.  Torrington Corp. has been trying to sell Ford 
  on roller cam bearings for years - the rollers would live fine in the 
  crankcase oil mist, no pressure feed required - but Ford evidently figures 
  it's cheaper to quietly swap whole motors rather than fix the problem or its 
  symptoms.  The symptom is journal wear; it can be cured by rollers.  The 
  problem is the long and torturous oil path from the sump, through the pump, 
  to the journals; it's a design flaw that would be harder to correct than 
  going to rollers. 

[email protected] (Dave Williams)
fangle  09 Nov 2000
- Most of the information I normally use on plain bearings came from 
  Machinery's Handbook, which has a sizeable section in the design of plain 
  bearings.  Rogowski's "Elements of the Internal Combustion Engine" is in 
  line with that, but it has some engine-related data: 
  - "The following data were taken on an eight-cylinder engine with a
     displacement of 221 in.^2, by motoring and firing at 2,000 RPM:
  - hp required to motor engine with pistons removed = 1.5
  - hp required to motor engine with pistons in place but cylinder
    head removed = 11.5
  - hp required to motor with engine completely assembled = 16.0"
- That's likely a flathead Ford V8.  The losses for "pistons removed", 
  assuming the rest of the engine was left alone, is much less than other 
  sources have indicated, while piston friction is much higher.  The usual 
  figure bandied about is 50% of the friction is in the pistons, the rest 
  scattered about the rest of the engine. 
- FYI, YMMV, eieio.

[email protected] (Dave Williams)
fangle  10 Nov 2000
- -> Was this with or without cam and lifters?
- I typed it in exactly as it was printed in the book.  "What you see is what 
  you get," as Flip Wilson used to say.  I wondered about the cam issue 
  myself... but the flatheads only used 60 or 70 pounds of spring pressure 
  when open; no need for lots when you just have a 1.5" valve, a retainer, a 
  hollow lifter, and 1/3 of spring weight. 

[email protected] (Dave Williams)
fangle  10 Nov 2000
- -> Wayne now mentions lubricant flow rate, bearing clearance, bearing
  -> load capacity - ah!!!, this was not found in my eleven months of
  -> intensive crankshaft research, using Dave Williams' and Robert
  -> Harris' 'usual sources'
- Did you ask?  Machinery's Handbook, Marks' and Kent's Handbooks, and of 
  course Rogowski, and (need I belabor the point) half a dozen papers at 
  naca.larc.nasa.gov. 
- All automotive sleeve bearings have such short aspect ratios they're running 
  mostly in the "edge effect" zone instead of a normal sleeve bearing.  Other 
  than that, any journal large enough not to allow excessive crank flex seems 
  to be fine. 

[email protected] (Dave Williams)
fangle  11 Nov 2000
- -> information nowhere else - and I have just near zero interest in 1920
  -> drillpress plain bearings  (is this wrong?).
- Wrong.  Plain bearings are plain bearings.  The information in Marks' and 
  Kent's is highly condensed but covers most applications, including internal 
  combustion engines.  You have to remember they're generally working with 
  different 'service factors' (same as electric motor ratings) than car guys 
  are used to; their recommended formulas will produce bearings the size and 
  shape of soup cans for a small block V8. 
- -> Elaborate on "edge effect", please?  (I'll be looking at those books,
  -> but it will take time to find them!).
- The rotating member floats off-center in the bearing on an oil wedge built 
  by its own rotation.  At the sides, it's easier for the oil to squirt out 
  the side than to maintain the wedge, so the edges carry very little load.  
  If a bearing is too narrow the bearing is depending mostly on the oil's 
  internal lubricity and its load rating goes way down. That's why a fully 
  grooved main bearing has less than half the load capacity of an ungrooved 
  bearing.